This paper compares the performances of the freeway traffic speed under the traditional weaving design and the new alternative promoted in recent years by both FHWA and state highway agencies. The new type of weaving segments features its addition of an exclusive lane for collection and distribution of on- and offramp flows at the interchange prior to their merges with the mainline traffic. Since the actual traffic condition at the interchange weaving areas may vary with not only its geometric features but also the volume distribution between the mainline and ramps, this study has developed a set of empirical models for analysis of the freeway mainline speed under the compound impact of geometry and traffic flows. Our analysis results with extensive microscopic simulation experiments have indicated that the new weaving design with a painted line to separate its C-D lane from the through lanes have resulted in a higher mainline speed under most traffic conditions. However, our simulation experiments have also revealed that the traditional weaving design may outperform the new weaving alternative under some congested traffic conditions, especially when heavy merging volumes have resulted in queue spillback on the through lanes.
Keywords: collector-distributor lane; weaving design; weaving performance evaluation